Kenya

1.Title of the survey:

Urban Labour Force Survey (ULFS), 1986.

Rural Labour Force Survey (RLFS), 1988/89.

2.Organization responsible for the survey:

Central Bureau of Statistics (CBS) and Long Range Planning Unit, Ministry of Planning and National Development.

3.Coverage of the survey:

(a) Geographical:

The 1986 ULFS covered urban areas and the 1988/89 RLFS, rural areas of the country. Both surveys excluded the northern half of Kenya.

(b) Persons covered:

All persons living in private households.

Excluded are institutional populations, non-resident citizens, foreigners and members of the armed forces.

In the 1986 ULFS, all persons between the ages of 15-64 were enumerated for potential membership in the labour force. Some information was also obtained for the 14 year olds and those above the age of 64, but published results only cover those aged 15 to 64 years.

In the 1988/89 RLFS, the economically active population relates to persons aged 8 to 64 years.

4.Periodicity of the survey:

The surveys are carried out on an ad hoc basis.

5.Reference period:

A single day, the day prior to the interview (known as the "yesterday" approach).

6.Topics covered:

The surveys provide information on employment, unemployment, underemployment, hours of work, income, duration of employment and duration of unemployment, discouraged workers, industry, occupation, status in employment and level of education.

7.Concepts and definitions:

(a) Employment:

The employed are persons of working age who, the day before the interview, did any work for pay, profit or family gain, or worked without pay on a farm or family holding (including unpaid family workers), and those who had a job and were temporarily absent from work.

Included are:

  1. full- and part-time workers seeking other work during the reference period;
  2. paid apprentices and trainees;
  3. participants in employment promotion schemes;
  4. paid and unpaid family workers, including those unpaid family workers who were temporarily absent from work;
  5. private domestic servants;
  6. seasonal workers awaiting agricultural or other seasonal work;
  7. persons on temporary or indefinite lay-off without pay;
  8. persons with a job but temporarily absent due to illness or injury; vacation or annual leave; maternity or parental leave; educational leave; absence without leave; bad weather; labour-management dispute; mechanical breakdown; or other reduction in economic activity.

Excluded from the employed are:

  1. persons who performed any work while being subject to compulsory schooling, or retired and receiving a pension;
  2. students;
  3. unpaid apprentices and trainees;
  4. persons engaged in own housework;
  5. members of producers' co-operatives;
  6. all members of the armed forces;
  7. persons doing unpaid community or social work.

(b) Underemployment:

The visibly underemployed comprise all persons in paid or self-employment involuntarily working less than the normal duration of work determined for that activity, who want another job or look for more work.

In the 1986 ULFS, a member of the labour force had to satisfy the following three conditions to be considered as visibly underemployed:

  1. He/she reported in the survey that he/she was unhappy with his/her current job because it was part-time;
  2. He/she must have looked for a job in the past week using a valid form of job search (see under the Unemployment definition); and
  3. The hours worked must be under the specified hours norm for his/her particular level of education and sex.

An attempt has also been made to estimate the significance of disguised underemployment, i.e. those individuals who are employed but whose earnings can be considered to be inadequate in relation to their training and work experience. For purposes of the ULFS, a wage norm was identified for each worker based on his/her level of educational attainment and used to estimate the degree of his/her disguised underemployment.

(c) Unemployment:

The unemployed are persons of working age who, the day before the interview, were without work and available for work and who have looked for work during the week prior to the interview. Also included are persons without a job and currently available for work who have made arrangements to start a new job on a date subsequent to the interview (no time limit is set for the new job to begin), and those persons waiting for a training.

"Looking for work" is interpreted as having taken one or more of the following actions: wrote to employer/relative, etc.; applied to Union Office or to Labour Office; answered newspaper advertisements; asked relatives, friends, etc.; approached employer directly, etc.

Full- or part-time students seeking full- or part-time work are excluded from the unemployed and considered as inactive.

Members of the population who are without work, available for work but not looking for work and who report that no work is available, are excluded from the unemployed and classified as "discouraged workers".

(d) Hours of work:

They refer to hours worked in all activities (main and secondary activities in the ULFS and farm and non-farm activities in the RLFS), the day and week prior to the interview.

(e) Informal sector:

This topic is not covered by the survey.

(f) Usual activity:

In both surveys, respondents are asked for their "usual" status in the last year, but data are not compiled on that basis.

8.Classifications used:

(a) Industry:

In the ULFS, the International Standard Industrial Classification of all Economic Activities (ISIC-1968) was used to classify employed persons only and coding was done to three digits. Published results are presented according to ISIC nine major divisions.

Information on the RLFS is not yet available.

(b) Occupation:

Data on the distribution of the employed are available from the 1986 ULFS, according to nine major groups which can be linked to the International Standard Classification of Occupations (ISCO-1968).

Data on the distribution of the unemployed by occupation profile of jobs being sought are also available from that survey, according to the same classification.

Information on the RLFS is not yet available.

(c) Status in employment:

In the 1986 ULFS, the fully employed and underemployed were categorized by job status into paid employees and self-employed. Paid employees include wage and salary earners (including those workers temporarily absent from work but retaining a formal attachment to their job), apprentices and trainees. Self-employed are sometimes further distinguished according to whether they are employers or not. Unpaid family workers are captured in the self-employed.

Since all respondents of working age were asked for all current job activities, but not previous activities, the unemployed were not classified by job status.

In the 1988/89 RLFS, data on status in employment of employed persons were collected according to the following status groups:

  1. Employee in private sector (whether paid or unpaid);
  2. Employee in public sector (whether paid or unpaid);
  3. Trainee or apprentice;
  4. Employee on family farm (whether paid or unpaid);
  5. Employee on non-family farm (whether paid or unpaid);
  6. Head of family farm;
  7. Head of non-farm income earning activity.

(d) Level of education/qualifications:

Information is collected on the highest education level of all respondents and classified according to eight groups:
  1. None,
  2. Standard 1 to 4 (primary education),
  3. Standard 5 to 8,
  4. Form 1 and 2 (secondary education),
  5. Form 3 and 4,
  6. Form 5 and 6,
  7. College,
  8. University.

Information is also collected on the highest certificate, degree or diploma obtained, according to nine codes. In addition, further information is collected and published on training received both from private and public sources, related to the current job and/or to the job sought.

9.Sample size and design:

(a) The sample frame:

It is based on the National Sample Survey and Evaluation Program (NASSEP), 1980/84, which was designed by the CBS and used in a number of different surveys in Kenya. The ULFS and RLFS are respectively based on the urban and rural portions of this frame.

The urban centers included in the sampling frame are defined as those which reported a population in excess of 2,000 in the 1979 Population Census and which were located in the area covered by the NASSEP sample. The sampled area excludes the northern half of Kenya. Since this area contains very few urban centers, over 95 per cent of the urban population in 1979 was covered by the 1986 ULFS.

(b) The sample:

In the 1986 ULFS, a stratified two-stage cluster design was adopted. Enumeration areas of the 1979 Census were selected as the primary sampling units (PSUs), using probability sampling proportionate to size based on the number of expected clusters of an average size of 100 households. Six large towns (Nairobi, Mombasa, Nakuru, Kisumu, Eldoret and Thika) were drawn into the sample. All other towns were grouped by provinces and selection of clusters from these centers was also made using probability proportionate to size. Within towns, stratification was carried out on the basis of income levels, in three categories: high, middle and low income zones. The complete sample consisted of 150 clusters.

Ultimate sampling units are households. Household listings were compiled by teams of field enumerators and households were selected from these lists by systematic random sampling to make the sample for the ULFS. All members of the selected households were covered by the survey. An up-to-date listing of households was obtained three months prior to field work.

The overall sample size was about 3,058 households, representing about 2 per cent of the urban population.

The corresponding information for the 1988/89 RLFS is not available yet.

(c) Rotation:

10.Field work:

(a) Data collection:

Data are collected by personal interviews carried out by field enumerators who are recruited on an ad hoc basis and trained by the CBS.

For the 1986 ULFS, data collection took place over six weeks, in November and December 1986. Each enumerator covered about two clusters, each containing between eight and 13 households. The 1988/89 RLFS was carried out in two phases: August-October 1988 and February-April 1989. Each Enumerator also covered about two clusters, and interviewed about 30 households.

(b) Substitution of ultimate sampling units:

When a sample household cannot be reached or where there is total non-response, no substitution is made.

11.Quality controls:

Extensive field supervision takes place throughout the field work. Questionnaires submitted by enumerators are checked for completeness by supervisors and revisits made to households for which errors or omissions are found. Supervisors also ensure that call-backs are made to maximize the response rate. Once a cluster is completed, the questionnaires are returned to district headquarters and then forwarded to CBS headquarters in Nairobi.

Data entry and analysis are made on microcomputers with built-in validity and consistency checks.

12.Weighting the sample:

Estimates of totals are made by weighting data for each individual respondent and summing up results over all respondents in the sample.

13.Sampling errors:

Not calculated.

14.Adjustments:

(a) Population not covered:

No adjustment is made.

(b) Under/overcoverage:

No adjustment is made.

(c) Non-response:

The 1986 ULFS experienced two types of non-response:
  1. Non-responding clusters - of the 150 clusters selected, questionnaires were returned from only 146;
  2. Non-responding households - the response rate was 88.2 per cent. Of the non-responding households, approximately 40 per cent were vacant structures, another 40 per cent represented households on leave, and 5 per cent were refusals. The remaining non-responding households were attributable to a variety of reasons.

The weights used to convert individual cluster returns to national urban totals were modified to reflect both forms of non-response.

No information is available on the 1988/89 RLFS.

15.Seasonal adjustment:

Because of the seasonality of activity in rural areas, the RLFS was conducted in two periods, August of 1988 and March of 1989 and the same households visited during each of these periods so that the results of the survey not be biased.

No other adjustments were made.

16.Non-sampling errors:

In the 1986 ULFS, enumerators reported occasional difficulties in obtaining information on personal incomes. A second minor problem occurred when the household head being questioned was different from the head listed during the generation of the sample. No other information is available.

17.History of the survey:

The first labour force survey was conducted in 1977/78 and covered both urban and rural areas of Kenya. The 1986 ULFS was restricted to urban areas and the 1988/89 RLFS complemented the ULFS.

The 1977/78 and 1986 surveys are based on slightly different definitions of labour force activity and unemployment. The most significant of these differences is the length of the reference period used in the survey.

The 1977/78 LFS used a single day reference period and cycled the survey over a one-year period. In the 1986 ULFS, data collection was limited to a six-week period and respondents were asked for their work activities "yesterday" as a single day reference period, but were also asked for their "usual" status. Job search activities were determined with respect to a one-week period, though activities yesterday were also determined to allow comparisons with the 1977/78 survey.

The 1988/89 RLFS used the "yesterday" and last week approach.

18.Documentation:

Central Bureau of Statistics: "Labour Force Survey, 1977/78, Basic Report" (Nairobi, Ministry of Planning and National Development, 1986).

idem: "Urban Labour Force Survey, 1986" (ibid., 1988).

The results of the 1988/89 RLFS are not yet available.

Non-published results can also be made available upon request.